INDEX

Explanations

periphery

The neuron consistently fires on mentions of “periphery,” i.e. references to outer boundaries or edges.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

AdvEx

-3.50

ⵈ

-3.27

釭

-2.94

腘

-2.89

犼

-2.73

 they

-2.70

鏷

-2.69

 infrastruktur

-2.64

塒

-2.64

fusions

-2.61

POSITIVE LOGITS

 They

3.42

Of

3.31

In

3.30

 mereka

3.22

 According

2.97

):

2.86

Of

2.84

2.77

ing

2.67

とはいえ

2.50

Activations Density 0.006%